Title:Growth-rate distributions of gut microbiota time series

Abstract:Logarithmic growth-rates are fundamental observables for describing ecological systems and the characterization of their distributions with analytical techniques can greatly improve their comprehension. Here a neutral model based on a stochastic differential equation with demographic noise, which presents a closed form for these distributions, is used to describe the population dynamics of microbiota. Results show that this model can successfully reproduce the log-growth rate distribution of the considered abundance time-series. More significantly, it predicts its temporal dependence, by reproducing its kurtosis evolution when the time lag $\tau$ is increased. Furthermore, its typical shape for large $\tau$ is assessed, verifying that the distribution variance does not diverge with $\tau$. The simulated processes generated by the calibrated stochastic equation and the analysis of each time-series, taken one by one, provided additional support for our approach. Alternatively, we tried to describe our dataset by using a logistic neutral model with an environmental stochastic term. Analytical and numerical results show that this model is not suited for describing the leptokurtic log-growth rates distribution found in our data. These results support an effective neutral model with demographic stochasticity for describing the considered microbiota.